Evaulating Datadog as a central metrics storage, conformance?

[Pix Pax]

Hya!

We are fairly invested in the Prometheus ecosystem, but from time to time, we are evaluating managed offerings as well.

Has anyone here had any experience with running Datadog at a scale that is powered by Prometheus instances?

We are using both the sidecar and the Receive pattern, and the tenant Prometheus servers are mostly Prometheus-Operators.

Our load is ~10M head series (depupped in Thanos, os 20M is the real data). 

Are you aware of any incompatibility maybe? I couldn't find DataDog in the conformance program.

Any feedback is appreciated!

Best,

Pax

[Stuart Clark]

We recently did an evaluation of Datadog (but ended up with going with
Grafana Cloud). In general Datadog seemed to be a good system, but like
a lot such offerings works best if you go "all in", rather than trying
to interoperate with other systems. As a result we didn't think it
worked for what we had/wanted (continuing with our existing Prometheus
servers instead of replacing everything with their agent and ingesting
Prometheus formatted metrics directly to Datadog).

What are you looking for that your current setup doesn't give you?

--
Stuart Clark